A few new ideas

[Fidd]

89. Listing, broken-keels and by the stem/stern sinkings.
It might be some nice eye-candy/interesting if the manner in which ships sank was related to the placement of the hit, and the type of ship. For example, a shallow hit might cause a ship to list and slow, a deeper hit a list, but faster sinking, a magnetic hit under the keel a possible "broken back" and so forth. As far as I recall, SH3 did this rather well. It may be the case this is done already, but as I'm very rarely on optics, I've never seen it!

[Fidd]

90. Merchants pitching, rolling, yawing in heavy seas, inversely proportional to weight for a given sea-state?

This might help complicate deriving range (from vertical measurements), AOB and convoy heading, so that a spectrum of difficulty is presented for games, helping to ensure player retention long-term, by providing scope for as much difficulty as old-sweats can handle.

[Fidd]

91. No safe depth from ASDIC/Hydrophones/DC's

I would like to see the end of the "get to 185m and you're completely safe from detection/DC attack".

To that end, there needs to be a new geometry applied to detection, with Asdic providing accurate range, bearing and depth, whilst the u-boat remains inside asdic range, and the escort has not closed to the range where the ASDIC loses contact. As the escort loses ASDIC contact, the escort hydrophones can estimate (rough) range and bearing, and speed of target (if heard) but not depth.

AI would be programmed to establish the depth, range and bearing, then to prosecute an attack at speed on the last known location from ASDIC, adjusting for hydrophone readings of range and bearing, but assuming no depth change.

Once ASDIC contact is lost, an escort (and any others joining the hunt) would execute a "lost contact" pre-programmed ASDIC search, which requires establishing some distance from the u-boat before commencement. If the u-boat is making noise, then that (bearing and rough-range) is used instead to bring the ASDIC onto the u-boat, resulting in a shorter interval for the next attack. If the boat has gone silent: dead-slow/creep-mode, no torpedo reloads (?) then it typically takes longer for an escort to re-establish ASDIC contact. The deeper a u-boat is, the greater the potential distance between where the escort drops DC's and the position and depth of the u-boat when they go off, again, if silent, the escort gains no adjustments from the hydrophones, and the DC'ing is less accurate.

If a u-boat manages to avoid being picked-up by AI ASDIC+Hydrophones, on it's lost contact search, then the escort prosecutes another. If the u-boat evades being located on the second, then it has escaped the escort and the convoy de-alerts. If the search period (configured in lobby) is medium, rather than short, then 3 searches have to be evaded, if long, then it's four.

The above system would greatly assist in forming a spectrum of difficulty for the game, by making operations in long period searches in shallow water extremely difficult to survive, through to short search periods in deep water being much easier to survive, but, and this is crucial, still involving some risk, as now ASDIC detections at 185m would still be possible, resulting in a DC attack, albeit a less accurate one...

[Fidd]

92. Shout accompanying the "Alarm bell".

Pressing the Alarm-bell causes a shouted pre-record "Alarm" from the bridge down into the tower and control room, if the boat is surfaced. If remote from the control-room , ie in the engine room, or if submerged, then just the alarm-bell is rung. NB the engine and emotor rooms need alarm bells! (but not switches) Consideration could be given to making the shouted "Alaaaarm" only occur if a player is actually on the bridge? The pre-recorded sound should be selected from a reasonable large number of exemplars, preferably of different German voices?

[Fidd]

93. Difficulty level bonuses.

Whilst it's possible to graduate through different roles in the boat, usually from helm towards captain, fundamentally any given role, once learned, confers no extra skills to be learned one the basic job is learned. This might be considered due to the need for bots to be able to do every role, or it might be that the devs thought that all tasks needed a basic accessability level for new players, rather than being a sim, it was a game.

This seems to me to be a missed opportunity in the first instance. Bots should be able to do all the basic tasks well enough, and players can learn the basics to a degree by watching them. However, if we take the dive-officer as an example, one could make extra valves operate for a player who wants more to do, or a greater challenge, by adding longitudinal trimming, necessary for depth keeping, or torpedo compensating tanks to deal with. If these controls did not operate when in use by the bots, or a newer player who does not wish to deal with those, then some roles on the boat could have their "life" extended by providing increased difficulty.

So why might crews want this? There's more risk if a DO is finding it harder to maintain a precise PD. Suppose that as you play on the harder settings - eg playing real radio rather than simple morse, using the Enigma etc, or in our DO's case, dealing with more realistic trim and compensating tanks, then that player could acrrue points for tonnage gained whilst playing the harder role, and lose it by getting killed. The SUM of the points aboard might permit one or two Lut, or Acoustic torpedoes in the pre-game loadout, or other "slightly early" adoption of kit for the NEXT game. It might be a higher capacity battery or similar. Having newer players on the boat would also add benefit in this way, to aid training and good experiences for new players, rather than getting the old sweats into a crew and then not helping bring on new players.

Possible voluntary extra difficulties:

Helm: realistic lag between rudder-movements and reactions of heading change.

Toppling of individual gyro-compass repeaters. Temporary complete loss of gyro-compasses (from DC attack).

Dive Officer: Compensating tanks for firing of torpedoes, longitudinal trim-tanks, lag in effect between any trim change and movement of the boat. Small lag in effect for planes and movement. Bouyancy instability if DC's detonate above, or below. Inertia for any given movement of the u-boat.

Radioman - mandatory encryption of all signals. (achieved by penalty of using words on a banned list for en-claire transmission. Real morse. HFDF of extended transmissions, or overly frequent ones, especially close to the convoy. Ability to operate DFing kit to plot a line towards another u-boat or escort, and a rough range. Faster transmission rate of bot-sent morse.

Engineers: More realistic recharge rates, need to move fuel to header tanks (once) per game. DC damage to exhaust valves causing diesel sea-water flooding (in addition to exhaust hull valves being left open when diving).

Captain: Need to sent and receive all signals encrypted. Otherwise all the optional difficulties from each role electing to play at a higher difficulty.

All: Torpedoes require manual player movement and alignment, insertions for reloading and (noisy) recovery of spares from below deck-plates, using winches etc. Manual firing of torpedoes in torpedo or emotor rooms. Need for torpedo doors to be opened and shut, heating and charging electric torpedoes before firing.

And so on. With a bit of thought, optional increased of difficulty could be added to most roles without causing problems for the bots provided that if a player with an optional extra difficulty on, comes off the controls, then those settings are reapplied when he returns, but are otherwise no longer occurring.

[Fidd]

94. Lights reacting to each other.

It would be good if when a uboat is detected, or the track of a steam torpedo is seen by any ship, if it's searchlight beam ceased moving around, and more or less fixed on that location (of the nearest detected uboat), with adjacent ships gradually similarly directing their search towards that point. Ideally, the more ships that do so, the more the detection-range of a surfaced uboat, or the periscope of one, increases. So instead of a detection range for the alert, and one for the unalarmed convoy, there'd be a gradual increase in detection range as more searchlights join in - if there's something to spot! A more nuanced spread of detection-ranges would help move the game into a more analogue "judging the risk" rather than the more "digital" "we're out of detection range at 2051m" sort of deal. It would also allow a lower detection range for the other side of the convoy, where fewer searchlights are trained his way....

[Fidd]

95. It further occurs to me that a similar "analogue detection ranges" could also be applied to ASDIC, so that if one uboat is detected on asdic, a number of nearby escorts join the hunt, with both the lateral and vertical depth of such searches being influenced by the number of escorts joining the hunt for a particular boat. This would likely make evading laterally increasingly more difficult, especially if the escorts joining the hunt are fast-moving. It would also complicate staying undetected, as it would no longer be a sure-fire evasion to CD to 185m. Instead, with one escort above you, the asdic "floor" might be 160m, with (say) 15m deeper per escort joining, so if there's 4 in all, then one would have to dive to 220m (gulp!) to be sure to evade Asdic. Randon crush-depths would ratchet-up the tension!

A nice side benefit of this is that attacking a convoy where the nearest escorts are slow-moving corvettes would be fairly safe, one with sloops and corvettes less so, and one with more destroyers than corvettes downright dangerous, because the escorts could combine more quickly than would occur if the nearby corvettes the only close escorts..... This would in effect, simulate (ish) the "creeping attack" where 1 escort maintained asdic contact, and joining escorts fired/released DC's on command from the stationary escort. This made it very difficult for the uboat to manoeuvre whilst the attacking escort was in the dead area for it's asdic. It proved fairly effective, and whilst the advent of "hedgehogs" and "squid" rendered it less necessary, it was used until the end of the war, as it was a more certain means of destroying a u-boat than simply having two escorts conducting individual attacks. This method would likely be a lot simpler to model in terms of AI behaviour, than formally coding AI cooperative hunts as movements?

EDIT - addition: Another nice aspect of this is that because the current certainty of safety at 185m is off the table, once a second escort joins the first, it's no longer possible to know when escorts have ceased attacking, meaning that bursts of speed and turns would serve to make re-detections more likely, and precipitate further attacks, with safety only being gained once at least one of the joined escorts have lost contact....